CleanCutAudio

①Introduction

What CleanCutAudio is and who this manual is for.

CleanCutAudio is an audio editor for Windows designed to clean up, polish and “master” recordings: voice, podcasts, music, samples. It works with MP3 and WAV files and offers three groups of tools: a waveform editor (cuts, fades, silence, reverse, pitch), a professional DSP effects chain (noise reduction, de-esser, equalizer, compressor, ambient, soft clipper, gain, stereo widener) and a set of analysis tools (LUFS/loudness, FFT spectrum analyzer, VU meters, mastering report).

This manual is meant to be read on two levels. The main text explains what each feature does and how to use it, in plain language. The blue-gray boxes marked ⚙️ Technical deep dive add details on the parameters (thresholds, ratio, LUFS, true peak…) for those who want to understand what happens “under the hood”. If you are a beginner you can safely skip the technical boxes: the rest of the text is all you need to get to work.

②The interface

How the main window is organized (v2.2.0.0).

At startup, until you load a file, the waveform shows a welcome screen inviting you to drag in an MP3 or WAV, plus a reminder of the main shortcuts. The window is divided into a few fixed areas:

The menu bar

Many features also have a quick button in the right-hand column or a keyboard shortcut. The DSP window (effects chain) and the Sound Analyzer open separately and can stay open while you work.

③Opening a file

Three ways to bring audio into the editor.

1. Drag and drop — take an MP3 or WAV file from the desktop or File Explorer and drop it onto the waveform.
2. From the built-in browser — move through the folders in the tree panel on the left, then double-click the file (or select it and press Enter, or right-click ▸ Load into Waveform).
3. From the menu — File ▸ 📂 Open… and pick the file in the dialog.

In the browser you can add the folders you use often to your Favorites (right-click ▸ Add to Favorites) and jump quickly to standard locations with the roots button. Once loaded, the file appears in the waveform and its data shows up in the information panel.

④Navigating the waveform

Zoom and scroll: moving precisely through the audio.

The waveform is your workspace. You move around with the mouse and keyboard:

The time ruler at the top shows the position in minutes and seconds. Hovering the mouse over the waveform highlights the regions; the cursor changes shape when you are near a draggable element (marker, playhead, fade handle).

⑤Selection, markers and regions

The heart of editing: marking out the portions you want to work on.

CleanCutAudio distinguishes two concepts that are best kept separate:

• Selection — the range between the start marker and the end marker. It is the portion used by loop playback, transcription, LUFS analysis of the area and commands such as “Trim to Selection”.
• Regions — portions of audio created by inserting cut markers. Each region has a name, a color, its own fades and — above all — its own independent DSP chain.

Start and end markers

Drag the two markers directly on the waveform to define the selection. During playback the selection stays consistent: if you move the markers, the playback position is brought back inside the range.

Cut markers and regions

1. Place the position where you want to split and press + to insert a cut marker (or Cuts/Regions ▸ Insert Marker Cut).
2. Each cut marker splits the audio into consecutive regions. Add as many as you need to isolate the parts (the tracks on one side of a cassette, the songs in a mix, the speakers in an interview…).
3. Remove a single cut with − (or Delete Marker Cut); remove them all with Ctrl+Shift+X (or Remove All Markers Cut).

Click a region to select it: it becomes the “active” region, the one the DSP window and several commands refer to. Press F10 to show/hide the list of regions and cuts.

✂️Marker Cut: the complete guide

The central concept of CleanCutAudio — where regions come from and everything that follows from them.

The cut marker (marker cut) is the element the program's entire workflow revolves around. It is not just a decorative “notch”: each marker cut divides the audio into regions and each region becomes a self-contained block, with its own fades, its own DSP chain and its own export behavior. Understanding marker cuts well means understanding how CleanCutAudio thinks.

What a marker cut is and what it creates

When you insert a marker cut, the audio to the left and to the right of the cut becomes two separate regions. Adding more gives you a sequence of consecutive regions covering the whole file. Each region carries:

Inserting, moving and removing marker cuts

The TIMELINE overlay (F10)

Pressing F10 brings up the TIMELINE overlay, the chronological map of the file: it lists, in order, MARKER START, each REGION with its range (e.g. Region 03 [1:46.237 → 2:27.690]), each CUT with its position and finally MARKER END. Regions are named automatically (Region 01, 02, …) and cuts stay “(unnamed)” until you rename them; the entry for the selected region is highlighted. It is the most accurate way to check cut points down to the millisecond.

The active region

Clicking a region makes it active: it is the region the DSP window (which shows the region's name and color at the top) and several editing commands refer to. In other words, to apply effects or actions to a specific portion, first you select it, then you act. Hovering the mouse over regions highlights them without activating them.

What follows from marker cuts

1 · DSP processing for different parts

This is the most powerful consequence: since each region can have an independent DSP chain, you can treat different parts of the same file differently. Typical examples: heavy noise cleanup only on the spoken intro; more compression on the chorus; brighter highs only on a poorly recorded verse; turning the de-esser off where it is not needed.

2 · Clean fades and transitions

Each region has its own fade-in/fade-out handles. When you drag the marker cut between two regions, the program prevents the fades of neighboring regions from overlapping beyond the allowed limit, so the join point stays free of clicks and gaps. It is the correct way to join two tracks or split a medley.

3 · Trim to region

With Cuts/Regions ▸ Trim to Active Region (or Edit ▸ Trim to Active Region) you discard everything else and keep only the selected region: useful for extracting a single song from a long recording split with markers.

4 · Multi-region export

On export, the summary lists each region with its range and its DSP chain. With no cuts, the single main region is exported (the range between the start/end markers); with several cuts, each region is rendered with its own settings.

Typical workflow with marker cuts

1. Load the long recording (e.g. one side of a cassette with several songs, or an interview).
2. Listen and press + at the start of each new part: you get one region per song/segment.
3. Fine-tune the cut points by dragging the markers; adjust the fades at the transitions.
4. Select one region at a time and set up its DSP chain (noise, EQ, compression…) in the F9 window.
5. Export: you get each region as a separate file, each with its own processing, or you trim to a single region with Trim to Active Region.

⑥Fades

Smooth fade-ins and fade-outs, with no clicks.

Each region (and the main region when there are no cuts) has two fade handles: one at the top left for the fade-in and one at the top right for the fade-out. Drag them toward the center to lengthen the fade, toward the edges to shorten it. The fade-in and fade-out durations are stored for each region.

Fades are used to avoid “clicks” at the ends of a cut, to blend two adjacent regions and to end a track gracefully. When you drag a cut marker between two regions, the program prevents the fades of neighboring regions from overlapping beyond the allowed limit, so the transition stays clean.

⑦Playback

Listening to your work, with or without effects.

• Play / Stop — the Space bar or the transport buttons. Playback starts from the playhead; if the playhead is outside the selection, it is moved back to the start of the selection.
• Loop — enable the Loop checkbox to repeat the selection continuously: useful for adjusting an effect “by ear”.
• Follow playhead — with this checkbox enabled, the view scrolls by itself to follow the playback position.
• Repositioning — drag the playhead on the waveform to move the listening point, even during playback.

⑧Edit menu: editing commands

All the operations that change the audio, from the Edit menu.

These commands act on the selection or on the active region, depending on the case. You can always go back with Undo.

⑨Pitch Shift

Raising or lowering the pitch (Edit ▸ Pitch Shift…).

Pitch Shift changes the pitch of the current region/selection. The window offers a slider in semitones, a Preview button (listen to a preview of about 2 seconds before applying) and the Apply and Cancel buttons.

1. Select the portion to transpose and open Edit ▸ Pitch Shift….
2. Set the shift with the semitone slider (negative values = lower, positive = higher).
3. Press Preview for a quick listen, then Apply to confirm.

⑩The DSP chain

The processing modules in series, with real-time preview.

Open the window with the DSP button, from the Processing ▸ RealTimeDSP… menu or with the F9 key. The DSP chain is the set of effects the audio passes through in sequence: the output of one module is the input of the next. Each region has its own chain, so the window always shows the parameters of the active region — the current region is shown at the top (e.g. Region: Main Region (MASTER)).

How the window is laid out

• Analyzer at the top — a real-time FFT spectrum (with a frequency/level tooltip on mouse hover) and a vectorscope showing the stereo image. The Preview DSP button turns listening with the effects applied on and off.
• Stereo Link Control — the Enable Equalizer Link L/R and Enable Compressor Link L/R checkboxes: when enabled, changes to one channel are automatically applied to the other.
• Module rack — the stack of effects, each with: an on/off switch (the small circle on the left), an arrow to expand/collapse the panel, and a preset drop-down. Note that Equalizer and Compressor are split into separate Left and Right panels.

Module order

The recommended order (“PRO order”), from top to bottom, is the order in which the signal is processed:

At the top of the list you will also find the “fixed” modules Stereo Widener and Stereo Linking (L/R channel linking), which act on the overall stereo image.

How to use a module

• Turn the module on/off with its switch: you can compare “before and after” on the fly.
• Preset — choose a starting point from the preset drop-down (each module has presets designed for typical cases).
• Parameters — adjust the knobs/values; with loop playback on, you hear the change immediately.
• Expand/collapse the module panel to make room; the open/closed state is remembered.

In the pages that follow, each module has a card describing what it is for, how to set it up, a table of its parameters with their ranges, and the factory presets.

Noise Reduction Module 1

Reduces constant background noise — hiss, hum, microphone noise — by tracking the noise level and lowering the signal when it drops below that threshold. It is the first link in the chain because it is best to clean up before equalizing and compressing.

How to set it up

Start from a preset (Light / Medium / Strong Clean) and loop a part with only noise and a part with the voice. Raise the Sensitivity until the background disappears, but stop as soon as the voice starts to sound “muffled” or to open and close unnaturally. The Reduction sets how deep the cut goes.

Factory presets

De-Esser Module 2

Softens sibilance — the “s”, “sh” and “z” sounds that become harsh and piercing, especially on close-miked voice. It acts only on the offending frequency band and only when that band exceeds a threshold.

How to set it up

1. Enable Solo Sibilants to hear only the band the de-esser is “watching”: move the Frequency until you clearly hear the “s” sounds.
2. Adjust the Bandwidth to narrow or widen that band around the sibilants.
3. Turn Solo off, then lower the Threshold until the “s” sounds soften, and use the Ratio to decide how much to reduce them.

Factory presets

Equalizer Module 3

Corrects the tonal balance: removes boominess from the lows, adds presence and “air” to the highs, carves out annoying frequencies. It is a 5-band parametric equalizer, with each band configurable independently.

How to set it up

For each band, choose the filter type, the frequency, the gain (cut or boost) and the Q (width of the adjustment). The frequency is adjusted on a logarithmic scale, as in professional plugins, so you get fine resolution in the lows.

Factory presets

The included presets are: LowEnd Cut (cleans up excess lows), Bass Boost, Treble Boost, Air Boost (opens up the highs), Voice Clean, Podcast Voice (low cut + presence on speech), Telephone Effect (“telephone” voice), HiFi Smile Curve (boosted lows and highs) and Mastering Gentle Curve (a gentle mastering smile curve). Each preset sets the 5 bands to ready-made values that you can then fine-tune.

Compressor Module 4

Evens out the dynamics: it reduces the gap between loud and soft parts, making the voice more consistent and present and the mix more “compact”. It is the module with the most parameters, with control over the knee, makeup and output ceiling.

How to set it up

1. Lower the Threshold until you hear the compressor working on the peaks.
2. Set a moderate Ratio (2:1–4:1 for voice) and adjust Attack/Release: a slower attack lets transients through, a suitable release avoids “pumping”.
3. Recover the lost volume with Makeup and protect the output with the Ceiling.

Factory presets

Ambient Effect

Adds a sense of space and air around the sound, like a small room/short reverb. Used sparingly it gives depth to the voice; pushed harder, it creates a “dreamy” effect.

Factory presets

SoftClipper Module 5

Adds warmth and “glue”, and at the same time acts as a soft limit on peaks: instead of clipping abruptly (hard clip), it rounds off the crests, gaining perceived loudness without unpleasant distortion.

Factory presets

Gain Module 6

The last link: it sets the final volume of the region and can apply a fade-in and fade-out at module level. This is where you set the final level before export or LUFS normalization.

Factory presets

Stereo Widener & Stereo Linking Stereo image

The Stereo Widener widens the stereo image, giving the sound a sense of spaciousness and “width”. Stereo Linking is the switch that links the Left and Right channels of the modules, so a change on one side applies to both.

Factory presets

⑪LUFS & loudness

Measuring and normalizing perceived loudness to the EBU R128 standard.

LUFS (Loudness Units Full Scale) measures how loud audio is to the human ear, not just technically. It is the yardstick streaming platforms use to match volumes: if your track is too loud, Spotify or YouTube turn it down; if it is too quiet, it falls behind the others. The LUFS Normalization window analyzes the file and tells you exactly where you are and how much gain you need to hit a target.

The window and its commands

Open it from the Normalize LUFS button or from Tools ▸ Loudness Normalization (LUFS)….

Target presets (LUFS)

Analyzing and normalizing

1. Press Analyze LUFS: the program measures the track (ITU-R BS.1770 mono downmix) and draws the graph.
2. Choose a target preset, or set Target LUFS and Ceiling by hand.
3. Read Gain to Target (how much would be needed) and Gain Allowed (how much is permitted without exceeding the ceiling): Gain APPLY is the smaller of the two. Listen with Preview LUFS, then Apply Normalization.

Batch normalization

At the bottom of the window, the NORMALIZE MULTIPLE FILES (LUFS) module normalizes several files at once with the same preset: add the files with Add Files… (or empty the list with Clear List), choose the output folder with Browse, then Start Batch (or Stop Batch). A Log and progress bars track the process. Handy for evening out the volume of an entire collection.

⑫Sound Analyzer

Measuring the quality of your track and comparing it with a reference track.

Open the Sound Analyzer from the button of the same name or from Processing ▸ Sound Analyzer…. It is not (just) a spectrum: it is a measurement station that, when you press Process And Analyze, calculates a set of indicators on the selected region and lets you compare them with a reference track that sounds “the way you want”.

The dynamics meters

Per-band analysis and comparison

Below the meters, the Loaded File Analysis panel breaks the track down into bands — Sub, Bass, LowMid, Mid, HighMid, High — and reports Fast LUFS, PLR, Low-Level Energy and Peak Sample. The Reference panel next to it shows the same values for the reference track, so you can see at a glance where your mix has too much or too little.

The FFT spectrum

The actual spectrum analyzer (20 Hz → 20 kHz, with a frequency/level tooltip) is built into the top of the DSP window, together with the stereo vectorscope. Its typical parameters:

⑬Mastering Report

A text report that sums up the analysis of the track.

The Mastering Report gathers the technical information and the suggestions about the current track in a single screen, organized into sections:

The report's Export button copies the entire text to the clipboard, ready to paste into a document or an email.

⑭Exporting

Saving the result with the DSP chain applied.

Exporting produces the final file with all the processing applied (each region's DSP, fades, any normalization). Before you proceed, the summary window (Export Summary) shows what will be exported: for each region its range and active DSP chain, or — if there are no cuts — the whole main region.

1. File ▸ 💾 Save As… (or the export flow) opens the summary.
2. Check the format, mode and bitrate; these values have defaults in the settings.
3. If needed, open the ID3 editor from the summary to fill in the tags before saving.
4. Confirm the export: a progress bar tracks the process.

⑮ID3 tag editor

The MP3 metadata: title, artist, album, cover art.

From File ▸ ✏️ Edit mp3 Info… (or from the export summary) you open the ID3 tag editor. Here you fill in the information that music players display: title, artist, album, year, genre and — depending on the build — the cover art. Filling in the tags is especially useful for podcasts and collections, where files are organized and displayed based on this data.

⑯Projects

Saving your work to pick it up later.

A project stores not only the audio but also your editing work: markers, regions, fades and the DSP chains of each region. That way you can close and pick up exactly where you left off.

⑰System recorder

Capturing the audio coming out of your computer (Tools ▸ Recorder…).

The recorder captures system audio — what you hear from your speakers: a live stream, a video, an online radio station — through the sound card's loopback. Press Record to start and Stop to finish; a level meter shows the incoming signal.

⑱Signal generator

Test tones and noise (Tools ▸ Tone Generator…).

The generator creates test signals, useful for calibrating a sound system, testing filters or creating reference material. Choose the signal type, then set the frequency, amplitude (in dB) and duration; for some types you can choose the stereo mode (Decorrelated, Phase Offset, Haas Delay).

⑲Voice & Text

Speech synthesis and transcription (Voice&Text menu).

Text-to-Speech (from text to voice)

With Voice&Text ▸ Text-to-Speech Module ▸ System Voice… you generate spoken audio from text, using the synthesis voices installed in Windows. If the computer has no usable voices, the program prompts you to install one from the Windows language settings.

Speech-to-Text (from voice to text)

With Voice&Text ▸ Speech-to-Text Module ▸ Convert Speech to Text you transcribe the portion between the start and end markers into text. The result appears in the Transcript window, with word timings when available.

⑳Audio settings

Choosing the driver and device (Audio Device menu) and the default values (Settings menu).

From the Audio Device menu you select the output driver and device. From the Settings menu, instead, you set the default values: Export Settings… (MP3/WAV format, CBR/VBR mode, bitrate) and DSP Settings… (startup presets and chain order). The choice of driver affects latency and compatibility:

⌨Keyboard shortcuts

The complete reference — keep it handy.

Playback and markers

Zoom and scrolling

Editing

Windows

📖Glossary

The technical terms you will come across in the program, in brief.

🛟Troubleshooting

Quick answers to the most common situations.

I can't hear anything during playback

In the Audio Device menu, check that the output device is the right one. If you chose WASAPI Exclusive or ASIO and another app is holding the device, switch to WASAPI Shared.

The system recorder doesn't capture anything

It is almost always WASAPI Exclusive mode: it prevents other apps from playing sound, so there is nothing to record. Set WASAPI Shared and try again.

Transcription says “model not found”

The Vosk model is missing: it must be in the Models/en folder next to the program. Download the language model and place it there.

Text-to-Speech has no voices

Install a voice from the Windows settings (Time & Language ▸ Speech). The program sees both SAPI5 and OneCore voices.

After normalization the volume did not go up as much as expected

The gain applied is the smaller of Gain to Target and Gain Allowed: if raising the level all the way to the target would exceed the ceiling, the program stops short so as not to distort. To go higher, you first need to reduce the peaks (compressor / soft clipper).

The voice sounds “muffled” after cleanup

Probably the Noise Reduction is too aggressive: lower Sensitivity and/or Reduction. Noise reduction should only be pushed until the background disappears, no further.

The file sounds good on headphones but weak in mono

Check the Stereo Widener: delay-based widening can weaken the signal in mono. Reduce the mix or turn it off for export if the material has to hold up in mono.

CleanCutAudio user manual — reference document generated from the application's features. The numeric values (parameter ranges and factory presets) reflect the program's current configuration and may change in later versions.